Labeling with heavy atom clusters attached to antibody fragments is an attractive technique for determining the 3D distribution of specific proteins in cells using electron tomography in the electron microscope. However, the small size of the labels makes them very difficult to detect by conventional bright-field electron tomography. We have developed a technique based on quantitative scanning transmission electron microscopy (STEM) and have demonstrated that it is possible to detect 11-gold atom clusters and 1.4 nm-diameter Nanogold particles, each containing an average of 67 Au atoms. STEM images of gold clusters embedded in carbon at a beam voltage of 300 kV were simulated using the NIST elastic scattering cross-section database. The calculated 2-D images of 11-gold atom clusters embedded in a homogeneous matrix indicate that the cluster visibility is maximized for low inner collection semi-angles of the STEM annular dark-field detector (15-20 mrad). [unreadable] [unreadable] Moreover, we have shown that STEM dark-field images collected at two different camera lengths yield quantitative distributions of both the heavy and light atoms in a stained biological specimen. Analysis of the paired STEM images reveals quantitative information about the distribution of fixative and stain within the biological matrix, and provides a basis for assessing detection limits for heavy-metal clusters used to label intracellular proteins. In sectioned cells that have been stained only with osmium tetroxide, we find an average of 1.2 +/- 0.1 Os atom per cubic nanometer, corresponding to an atomic ratio of Os:C atoms of approximately 0.02, which indicates that small heavy atom clusters of 11-gold and Nanogold can be detected in lightly stained specimens. Our calculations also indicate that the visibility of undecagold clusters in 3D reconstructions is significantly higher than in 2D images when the clusters are embedded in an inhomogeneous matrix corresponding to local fluctuations in composition. The experimental measurements show that it is possible to detect nanogold particles in plastic sections of tissue freeze-substituted in the presence of osmium. STEM tomography has the potential to localize specific proteins in permeabilized cells using antibody fragments tagged with small heavy atom clusters. Our quantitative analysis provides a framework for determining the detection limits and optimal experimental conditions for localizing these small clusters.[unreadable] [unreadable] We have also investigated the application of STEM tomography for determining the 3-D structure of thick (> 1 micrometer) sections of cells. Monte Carlo electron trajectory simulations have provided us with a method for comparing optimum strategies for signal collection in the STEM. To minimize geometrical beam broadening, it was important to use a small STEM probe convergence angle of approximately 1-2 mrad. It was also found that bright-field gives better spatial resolution than dark-field STEM throughout the volume of thick samples because electrons that have undergone multiple scattering emerge at higher scattering angles from the bottom surface. Experiments are in progress to use this approach to determine the 3-D structure of small whole eukaryotic cells, such as malaria parasites.